From owner-freebsd-current  Sat Nov 30 11:23:12 1996
Return-Path: owner-current
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id LAA02777
          for current-outgoing; Sat, 30 Nov 1996 11:23:12 -0800 (PST)
Received: (from wpaul@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id LAA02770;
          Sat, 30 Nov 1996 11:23:11 -0800 (PST)
From: Bill Paul <wpaul>
Message-Id: <199611301923.LAA02770@freefall.freebsd.org>
Subject: Re: exports rpc time out
To: asami@freebsd.org (Satoshi Asami)
Date: Sat, 30 Nov 1996 11:23:11 -0800 (PST)
Cc: current@freebsd.org
In-Reply-To: <199611300844.AAA00959@silvia.HIP.Berkeley.EDU> from "Satoshi Asami" at Nov 30, 96 00:44:47 am
X-Mailer: ELM [version 2.4 PL24]
Content-Type: text
Sender: owner-current@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> 
> I'm not sure if this is an NIS/YP problem, but here it goes.  When I
> have a long export list, "showmount -e" doesn't work:

I don't think this is an NIS problem per se. One way to test it is
to dump out your netgroups map into /etc/netgroup and use that (i.e.
the raw file) rather than NIS netgroups. If /etc/netgroup exists,
has real data in it and no '+', then getnetgrent(3) won't use YP.
If the problem persists, then the problem is probably not due to YP.

> ===
> ## showmount -e 
> RPC: Timed outCan't do Exports rpc
> ===
> 
> /etc/exports isn't that big but it uses some netgroups, each of which
> expands to a few dozen hostnames.  The exporting itself seems to be
> working ok, it's just that I can't see the list with showmount.

Hm... if the exporting works okay, the the initial netgroup lookups
must be working. I have to look at mountd to see how it figures out
what to tell showmount when you request a mount list.

> This is the same loaded net that was giving me all the sig11's before
> (which is fixed now, thanks) due to YP timeouts.

One think I notice is that showmount uses callrpc(), which I think
sets the timeout on the RPC call to 25 seconds. If mountd takes longer
than this to respond, the request will time out. This would mean that
mountd is taking too long to put together its response; this _might_
be due to NIS lookups taking too long, but I need to analyze the code
to be sure.

It might be possible to just increase the timeout.

-Bill